WORLD HEALTH ORGANIZATION REGIONAL OFFICE FOR THE WESTERN PACIFIC
Improving
DATA QUALITY
DATA QUALITY
A GUIDE FOR DEVELOPING COUNTRIES
IMPROVING DA
TA
QU
ALIT
Y: A GUIDE FOR DEVEL
OPING C
OUNTRIES
WORLD HEALTH ORGANIZATION
Regional Office for the Western Pacific United Nations Avenue
1000 Manila, Philippines Fax No : (63-2) 521-1036 Tel. No : (63-2) 528 8001 Email : [email protected] Website: http://www.wpro.who.int ISBN 92 9061 050 6
WORLD HEALTH ORGANIZATION
REGIONAL OFFICE FOR THE WESTERN PACIFIC
A GUIDE FOR DEVELOPING COUNTRIES
ISBN 92 9061 050 6 (NLM Classification: WA 950)
© World Health Organization 2003 All rights reserved.
The designations employed and the presentation of the material in this publication do not imply the expression of any opinion whatsoever on the part of the World Health Organization concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. Dotted lines on maps represent approximate border lines for which there may not yet be full agreement.
The mention of specific companies or of certain manufacturers’ products does not imply that they are endorsed or recommended by the World Health Organization in preference to others of a similar nature that are not mentioned. Errors and omissions excepted, the names of proprietary products are distinguished by initial capital letters.
The World Health Organization does not warrant that the information contained in this publication is complete and correct and shall not be liable for any damages incurred as a result of its use.
Publications of the World Health Organization can be obtained from Marketing and Dissemination, World Health Organization, 20 Avenue Appia, 1211 Geneva 27, Switzerland (tel: +41 22 791 2476; fax: +41 22 791 4857; email: [email protected]). Requests for permission to reproduce WHO publications, in part or in whole, or to translate them – whether for sale or for noncommercial distribution – should be addressed to Publications, at the above address (fax: +41 22 791 4806; email: [email protected]). For WHO Western Pacific Regional Publications, request for permission to reproduce should be addressed to Publications Office, World Health Organization, Regional Office for the Western Pacific, P.O. Box 2932, 1000, Manila, Philippines, Fax. No. (632 521-1036, email: [email protected]).
2
1
Contents
Acknowledgement
vi
Introduction
1
Aims and objectives of these guidelines 2
Target readership 2
Using the guidelines 2
The book’s structure 2
Defining data and information
4
What are data? 4
Primary and secondary data 5
Data and information 5
Health care data 5
Health Information 6
Health care information 6
Importance of data quality 7
Summary 8
References 9
Data quality
10
Data quality — what it means 10
Importance of data quality in health care 11
Components of data quality 11
Accuracy and validity 11
iv
4
3
Completeness 12 Legibility 13 Timeliness 13 Accessibility 13Development and implementation of standards for health care documentation 14
What are standards and how are they developed? 14
Clinical data standards 14
Benchmarking 17
Leadership in data quality 18
Routine data quality monitoring 18
Quantitative analysis of medical records 18
Qualitative analysis of medical records 19
Data entry checks 20
Checks on the quality of abstracted data 20
Audit of accessibility of medical/health records 21
Medical/health record audit 21
Development of an on-going quality assessment plan 22
Performance improvement techniques 23
Other steps to assist with data quality improvement 24
Limitations in overcoming problems related to data quality 25
Summary 25
References 26
Quality in data collection
28
Type of data collected and how collected 29
Administrative data 29
Clinical data 30
Collection of data in specialist outpatient clinics 31
Patient-held health record 32
Hospital census data 32
Secondary health care data collection 33
Causes and sources of poor data collection 33
Examples of how inaccuracies in data collection arise 34
Ways to overcome problems of poor data collection 35
Review of data collection forms 35
Staff requirements and training 35
Standards and checks 36
Why some data are collected and not used by planners, managers, etc. 37
Summary 37
References 37
Clinical coding quality
38
Defining clinical coding 38
Importance of accurate and timely coded clinical data 39
v
6
5
Analysis of the medical record 41
Selection of main condition 41
Allocation of correct codes 42
What to code? 42
Mortality coding 43
Importance of coding quality 43
Causes and sources of poor data quality in clinical coding 45
Methods that could be implemented to improve data quality in clinical coding 47
Development of clinical coding standards 47
Development of coding quality indicators 48
Other methods for improving data quality in clinical coding 49
Samples of studies in clinical coding quality 51
Summary 52
References 53
Data quality of statistical reports
54
Health care statistics 55
Standardization of statistical terms and formulae 55
Determining the appropriateness of statistical reports 56
Data presentation 57
Tables 57
Graphs 58
Computer generated statistics 59
Data quality of vital statistics 59
Births 59
Deaths 60
Causes and sources of poor quality statistical data 60
Methods to improve data quality of statistical reports 61
Other methods to improve the quality of statistical reports 62
Summary 63
References 63
Data quality in public health records and reports
64
Data collection and dissemination 64
Public health services 65
Causes and sources of poor data quality 66
Methods to improve data quality in public health data collection 66
Summary 67
References 67
Glossary
68
Appendices
72
Appendix 1: Medical Record Review Form 72
Appendix 2: Simple Data Collection Form for a Manual Medical/Health Record Tracking System 73
vi
Acknowledgement
WHO Regional Office for the Western Pacific acknowledges the contributions made by
Professor Phyllis J. Watson, Head of School of Health Information Management
(formerly), Faculty of Health Sciences, University of Sydney, to this publication.
1
Introduction
Accurate, timely and accessible health care data play a vital role in the
planning, development and maintenance of health care services. Quality
improvement and the timely dissemination of quality data are essential if
health authorities wish to maintain health care at an optimal level.
In recent years, data quality has become an important issue, not only because of its importance in promoting high standards of patient care, but also because of its impact on government budgets for the maintenance of health services.
Authorities at all levels of health care, including hospitals, community health centres, outlying clinics and aid posts, as well as ministries or departments of health, should be concerned about poor data quality and the impact it has on the quality of health care. In many countries, administrators are dogged by poor medical/health record documentation, large backlogs of medical records waiting to be coded and inconsistent coding, plus poor access to, and utilization of, accurate and accessible morbidity data.
These concerns not only relate to the quality of medical/health record documentation but also to the collection of health care statistics at all levels, from the largest hospital to the smallest clinic or aid post. At the hospital or clinic level, statistics are used to assess how much services are being used to enable the facility to make appropriate financial and administrative plans, and to conduct vital research. At state, province and national levels, governments use health statistics for planning health care services and allocating resources where they are needed most. The accuracy and relevance of the information processed, thus, are vital to the smooth running of the facility and also in assisting governments with decisions on the provision of health care services locally and nationally.
I
n recent years,
data quality
has become an
important issue,
not only because
of its importance
in promoting
high standards of
patient care, but
also because of
its impact on
government
budgets for the
maintenance of
health services.
2
Many clinics and aid posts offer immunization programmes, counselling, family planning and other health-related services that are not always considered parts of “traditional” health care but are, in fact, important components within the community. As such, they must maintain accurate and reliable data.
To address these issues and improve the quality of data collected, as well as the information generated from that data, quality-control measures need to be taken.
Aims and objectives of these guidelines
The purpose of this booklet is to provide a set of guidelines to enable health care workers, health information managers and administrators at all levels to focus on improving the timeliness, accuracy and reliability of health care data. Although the emphasis might seem to be on hospitals and hospital medical records, these guidelines have been designed to address all areas in health care where data are collected and information generated. The guidelines describe the activities that should be considered when addressing the question of data quality in health care, regardless of the setting. The reader is guided to assess and, where necessary, improve the quality of data generated in the environment within which they function, regardless of size, remoteness or sophistication.Target readership
These guidelines are aimed at government policymakers and health care administrators at primary, secondary and tertiary levels of health care, as well as doctors, nurses, other health care providers and health information managers. All of these share responsibility for the documentation, implementation, development, and management of health information services.
Using the guidelines
It is important that the reader treats these guidelines not as a set of definitive rules applicable in every situation but as an aide memoire, to help ensure that important activities have been considered and addressed to improve data and ensure data quality.
Although the term “patient” is generally used in the following pages, it is important to note that health services in many countries also use terms such as “consumer” or “client.”
The book’s structure
The text has been divided into six chapters. The first two chapters deal with defining data and data quality. In Chapter 1, data are defined and the difference between data and information explained. In Chapter 2, data quality is defined along with a description of the components of data quality. Data
3
standards and their implementation are also outlined together with quality control methods and performance improvement measures. Limitations in overcoming problems related to data quality are also discussed along with a description of some measures that can be used to solve identified problems of data quality.
Chapters 3, 4, 5 and 6 deal with specific areas of data quality. Chapter 3 deals with the type of data collected, how they are collected, causes and sources of poor data collection, and methods that could be used to overcome such problems.
Chapter 4 is dedicated to clinical coding quality and, again, issues relating to the causes and sources of poor quality data and methods to improve the quality of coded clinical information are highlighted. An outline of activities that could be used to measure coding productivity and conduct coding audits is included.
In Chapter 5, special attention is given to the quality of statistical reports, including the need for standardization of terms and formulae and the appropriateness and presentation of statistical information. The causes and sources of poor quality statistical data and methods to improve the quality of statistical reports are also covered.
Last, Chapter 6 deals with causes and sources of poor data quality in public health records and reports, along with possible methods to improve the quality of public health data collections.
4
Defining data
and information
1
The starting points for health care information are data and the collection of
data, whether maintained manually or electronically at a large teaching
hospital, health centre or outlying clinic. Demographic and clinical data stored
in a patient’s medical/health record are the major source of health information
and are of no value to medical science or health care management if they are
not accurate, reliable and accessible.
The comparison of health care data between facilities, states or provinces, within a country or between countries, is vital to the growth and dissemination of health information throughout the world. This possible sharing is meaningless, however, without the use of standardized systems for data collection, disease classification and health care statistics.
Before discussing data quality, we need to define data, how data are transformed into information and how they are collected.
What are data?
Data may be defined as a representation of facts or concepts or instructions in a formalized manner, suitable for communication, interpretation or processing by manual or electronic means. An element of data is an item, idea, concept or raw fact (Abdelhak et al., 1996).
In health care, these facts describe specific characteristics of individual patients. Whether we collect data on paper or in a computer, the data should be organized in such a way that we can understand and retrieve them when needed (Davis and LaCour, 2002).
W
hether we
collect
data on paper or
in a computer,
the data should
be organized in
such a way
that we can
understand and
retrieve them
when needed.
5
Primary data are obtained from the original data
source. That is, documentation in the patient’s
medical/health record collected by staff at either a
hospital, clinic or aid post. Daily ward census reports
collected in hospitals are also primary data.
Secondary data are data sets derived from primary
data. Secondary data are individual or aggregate
health care data found in reports that are summarized
from the source. At the hospital or health centre level,
secondary data include the master patients’ index,
disease and procedure indexes, health care statistics and
disease registries. At primary care level, they also
include such aspects as the patients’ name index and
statistics.
Data and information
In order to interpret data – to make sense of the facts and use them – the data items need to be organized. Once organized, data become information.
Health care data
Health care data are items of knowledge about an individual patient or a group of patients. In health care, data are captured about a patient in paper or electronic format during his or her attendance at an outpatient clinic, community health centre, primary health care provider, or his or her admission to a hospital (Davis and LaCour, 2002). The data collected should include all relevant findings relating to the patient’s condition, diagnoses, treatment, if any, and other events as they occur.
Whether the data are collected manually or in a computer, it is important to ensure that the information is correct at the point of entry.
Primary and secondary data
There are two types of data – primary data and secondary data. In the context of health care, primary data and secondary data are distinguishable as follows (Abdelhak et al., 1996).
6
To ensure data quality, two key principles are data accuracy and data validity. To communicate effectively, data must be valid and conform to an expected range of values. To be useful, data must be accurate.
As the recording of data is subject to human error, there needs to be built-in control measures
to eliminate errors, both in manual recording and computer entry.
Health Information
Health information is health care data that have been organized into a meaningful format.
Health information may refer to organized data collected about an individual patient, or
the summary of information about that patient’s entire experience with his or her health care provider (Davis and LaCour, 2002).
Health information can also be the aggregate information about all patients that have attended or been admitted to a hospital, or attended a health centre, outlying clinic or a community immunization or health screening programme.
Health information, therefore, can encompass the organization of a limitless array and combination of possible data items (Davis and LaCour, 2002).
Providers of health care services need information
not only at the point of service but also at the point
of decision making in a format that maximizes
the decision-making process.
Health care information
As a by-product of a patient’s contact with a health care provider, regardless of the setting, information should have value as a clinical review or management tool. Whether in a manual system or computer, information will not be valuable unless it is accurate, relevant, structured and presented in an easily useable form. Health care information should be capable of:
■ promoting excellent clinical care;
■ describing the types of individuals using services and the types of services they receive;
■ measuring efficiency of the contact, treatment, referral and interaction by health care professionals;
■ helping in the co-ordination of care between services provided;
■ providing meaningful statistics for determining the health status of the community;
■ measuring quality from a patient and provider perspective; and
7
Accurate information about resources used and services delivered at all levels of health care is essential for resource management, use of clinical evidence and measurement of outcomes to improve the effectiveness of health care services.
Correct and up-to-date information is critical, not
only for the provision of high-quality clinical care, but
also for continuing health care, maintaining health care
at an optimal level, clinical and health service research,
and planning and management of health systems.
The collection and use of information should not impose
a burden on the health system. It should be collected as
a routine by-product of the health care process.
Importance of data quality
Accurate and reliable health care data are needed for:■ determining the continuing and future care of a patient at all levels of health care;
■ medico-legal purposes for the patient, the doctor and the health care service;
■ maintaining accurate and reliable information about diseases treated and surgical procedures performed in a hospital and within a community, as well as immunization and screening programmes, including the number and type of participants;
■ clinical and health service research and outcomes of health care intervention, if required;
■ accurate, reliable and complete statistical information about the uses of health care services within a community;
■ teaching health care professionals; and
■ working out staffing requirements and planning health care services. Accurate and reliable health care data are used by:
■ doctors, nurses and other health care professionals treating a patient admitted to a hospital or seen in an outpatient department or emergency room, and in community health centres, outlying clinics or general practitioners’ offices. They use the medical/health record as a means of communication during an episode of care and treatment of a patient, and as an aide memoire for continuing care of that particular patient. Doctors also use health care data to evaluate the services provided;
■ nursing staff in hospitals to evaluate data and develop critical pathways and patient care plans for admitted patients;
8
■ health insurers who require information to reimburse the patient and/or health care facility for services rendered whether for an inpatient or ambulatory patient;
■ legal representatives and courts as documentary evidence of a patient’s care and treatment by a health care worker in a hospital, health centre or clinic. They are also used to protect the legal interests of the patient, doctor and other health care professionals, the health care facility, and the public;
■ the state and national ministry of health to review vital statistics and the incidence and prevalence of disease in a city, state or country. The provision of accurate and reliable aggregate data is important for public policy development and funding of health care services;
■ quality assurance committeesand medical staff as a basis for analysis, study and evaluation of the quality of health care services rendered to patients;
■ researchers, to analyse and interpret data to determine causes, prevention methods and treatment for diseases, injuries and disabilities;
■ health care facility accreditation and licensing agencies to review medical/health records to provide public assurance that quality health care is provided; and
■ national governments who use the information to develop health care policy and provide and regulate funds.
If data are not correct and valid, the usefulness for the
above purposes will be jeopardized.
Summary
To maintain data quality, health care data must be (Abdelhak et al., 1996):
■ appropriate – for the overall needs of the patient and health care services;
■ organized –to enable it to be understood and used when required;
■ timely and available – to users when required;
■ accurate and complete – otherwise it could be difficult to make appropriate decisions; and
■ Cost-effective – if the cost of collecting and disseminating information exceeds its value, the usefulness of the data collected must be addressed.
In the following chapters, the importance of data quality will be discussed in more depth along with guidelines on how to maintain quality in all areas of documentation, including the collection of data at the original source. Data quality in clinical coding, statistical collections and reports, as well as the dissemination of information to health care authorities, will be covered.
9
References
Abdelhak M, Grostick S, Hankin MA, Jacobs E. Health Information: Management of a Strategic Resource.
Philadelphia, WB Saunders Company, 1996.
Davis N, LaCour M. Introduction to Information Technology. Philadelphia, WB Saunders Company, 2002. Johns ML. Health Information Management Technology. Chicago, American Health Information
10
Data quality
2
As coded health care data are being increasingly used in the health care
environment, it is important to ensure that the
original source data
are
accurate and timely, which in turn, will produce reliable and useful information.
Regardless of whether in a hospital, health centre or an outlying clinic, the quality of health care data and statistical reports has come under intensive scrutiny in recent years. Thus, all health care service employees, including clerical staff, health professionals, administrators, and health information managers, need to gain a thorough knowledge and understanding of the key components of data quality and the requirements for continuous data improvement.
Data quality — what it means
In general terms, quality, as defined by Donabedian (1988), consists of the ability to achieve desirable objectives using legitimate means. Quality data represent what was intended or defined by their official source, are objective, unbiased and comply with known standards (Abdelhak et al., 1996). Data quality includes:
■ accuracy and validity –of the original source data;
■ reliability – data are consistent and information generated is understandable;
■ completeness – all required data are present;
■ legibility – data are readable;
■ currency andtimeliness – data are recorded at the time of observation;
■ accessibility – data are available to authorized persons when and where needed;
■ meaning or usefulness – information is pertinent and useful; and
T
he original
data must be
accurate in order
to be useful. If
data are not
accurate, then
wrong
impressions and
information are
being conveyed
to the user.
11
■ confidentiality and security – both important particularly to the patient and in legal matters.
The central theme is that data quality is proportionate
to the attainment of achievable improvements in health
care.
Importance of data quality in health care
Health care data are maintained for the present and future care of the patient regardless of the level at which the service is provided. The quality of that data is crucial, not only for use in patient care, but also for monitoring the performance of the health service and employees. Data collected and presented must be accurate, complete, reliable, legible and accessible to authorized users if they are to meet the requirements of the patient, doctor and other health professionals, the health care facility, legal authorities, plus state, province and national government health authorities.
Components of data quality
Accuracy and validity
The original data must be accurate in order to be useful. If data are not accurate, then wrong impressions and information are being conveyed to the user (Davis and LaCour, 2002). Documentation should reflect the event as it actually happened. Recording data is subject to human error and steps must be taken to ensure that errors do not occur or, if they do occur, are picked up immediately.
Example of accuracy and validity in a manual medical record system
■ The patient’s identification details are correct and uniquely identify the patient.
■ All relevant facts pertaining to the episode of care are accurately recorded.
■ All pages in the health record are for the same patient.
■ The patient’s address on the record is what the patient says it is.
■ Documentation of clinical services in a hospital is of an acceptable predetermined value.
■ The vital signs are what were originally recorded and are within acceptable value parameters, which have been predetermined and the entry meets this value.
■ The abstracted data for indices, statistics and registries meet national and international standards and have been verified for accuracy.
■ The codes used in hospitals to classify diseases and procedures conform to pre-determined coding standards.
In a manual system, processes need to be in place to monitor data entry and collection to ensure quality. In a computerized system, a computer can be instructed to check specific fields for validity and
12
alert the user to a potential data collection error. Computer systems have in-built checks such as edit and validation checks, which are developed to ensure that the data added to the record are valid. Edits or rules should be developed for data format and reasonableness, entailing conditions that must be satisfied for the data to be added to the database, along with a message that will be displayed if the data entry does not satisfy the condition. In some instances, the computer does not allow an entry to be added if it fails the edit. In other instances, a warning is provided for the data entry operator to verify the accuracy of the information before entry.
Examples of edits and validity in a computer-based system
■ In a hospital system, a patient must have a unique number because it is the key indexing or sorting field.
■ The patient’s number must fall within a certain range of numbers or the computer does not allow the data entry operator to move to the next field or to save the data.
■ For hospital patients, the date of admission must be the same as or earlier than the date of discharge.
■ A laboratory value must fall within a certain range of numbers or a validity check must be carried out.
■ Format requirements such as the use of hyphens, dashes or leading zeros must be followed.
■ Consistency edits can be developed to compare fields – for example a male patient cannot receive a pregnancy test.
Reliability
Data should yield the same results on repeated collection, processing, storing and display of information. That is, data should be consistent.
Examples of reliability
■ The diagnosis recorded on the front sheet of the hospital medical record is consistent with the diagnosis recorded in the progress notes and other relevant parts of the medical record.
■ Surgical procedures recorded on the front sheet of the hospital medical record are the same as recorded in operation reports in the body of the medical record.
■ The age of the patient recorded on the first sheet of a medical/health record is the same as that recorded on other pages.
■ The correct name of the patient is recorded on all forms within the medical/health record at the point of care or service given.
Completeness
All required data should be present and the medical/health record should contain all pertinent documents with complete and appropriate documentation.
13 Examples of completeness
■ The first sheet of the medical/health record contains all the necessary identifying data to uniquely identify an individual patient.
■ For inpatients, the medical record contains an accurately recorded main condition and other relevant diagnoses and procedures and the attending doctor’s signature.
■ Also for inpatients, all progress notes — from date of admission to discharge or death — are complete with signatures and date of entry.
■ Nursing notes, including nursing plan, progress notes, blood pressure, temperature and other charts are complete with signatures and date of entry.
■ For all medical/health records, relevant forms are complete, with signatures and date of attendance.
Legibility
All data whether written, transcribed and/or printed should be readable.
Examples of legibility
■ Handwritten demographic data are clearly written and readable.
■ Handwritten notes are clear, concise, readable and understandable.
■ In all medical/health records, undecipherable codes or symbols are not used in either manual or electronic patient records.
■ If abbreviations are used, they are standard and understood by all health care professionals involved in the service being provided to the patient.
Timeliness
Information, especially clinical information, should be documented as an event occurs, treatment is performed or results noted. Delaying documentation could cause information to be omitted and errors recorded.
Example of timeliness
■ A patient’s identifying information is recorded at the time of first attendance and is readily available to identify the patient at any given time.
■ The patient’s past medical history, a history of the present illness/problem as detailed by the patient, and results of physical examination, is recorded at the first attendance at a clinic or admission to hospital.
■ On discharge or death of a patient in hospital, his or her medical records are processed and completed, coded and indexed within a specified time frame.
■ Statistical reports are ready within a specified time frame, having been checked and verified.
Accessibility
Allnecessary data are available when needed for patient care and for all other official purposes.The value of accurately recorded data is lost if it is not accessible.
14
Examples of accessibility
■ Medical/health records are available when and where needed at all times.
■ Abstracted data are available for review when and where needed.
■ In an electronic patient record system, clinical information is readily available when needed.
■ Statistical reports are accessible when required for patient-care committees, planning meetings and government requirements.
Development and implementation of
standards for health care documentation
The primary purpose of recording data is communication. To make sure that communication is achievable, it is important to ensure that the language of health information, regardless of the level of service and information concepts, conform to a recognized standard.
Prior to introducing procedures to monitor data quality, the health care facility, whether a hospital, health centre, or outlying clinic, should look at implementing standards for health care documentation, processing and maintenance.
What are standards and how are they developed?
A standard is a specimen or specification by which
something may be tested or measured.
In health care, standards are defined as:
Best practice principles and guidelines for collection and
storage of health care data
(Abdelhak et al., 1996).
Standards that define demographic and other identifying data elements suited for data capture and use in patient identification in health care settings and provide guidelines for their application should be developed and adhered to by staff.
Clinical data standards
Clinical data standards should also be developed to ensure that data collected in one facility means the same in another, allowing for comparison between facilities. Two or more persons looking at the content of a medical/health record may have different opinions about its adequacy. An assessment about the adequacy of patient care information in the absence of any explicit standards is a value judgement.
To be useful, standards should be developed to promote
data quality at the time of documentation.
In many countries, to achieve standardization of data, steps such as the development of
15 Minimum Data Set
A minimum data set is a core set of data elements agreed to by a National Health Information Group for the mandatory collection and reporting at a national level.
It is reliant on a national agreement to collect uniform data and to supply it as part of the national collection, but does not preclude a health care facility from collecting additional data to meet its own needs.
Australia and the United States of America are two countries that have national minimum data sets. These are used to identify data elements that should be collected by health care facilities for each patient and provide uniform definitions for common terms. In the United States of America, the list of patient-specific data items used is the Uniform Hospital Discharge Data Set (UHDDS) (Johns, 2002).
Data element Definition/Descriptor
01. Personal identification The unique number assigned to each patient within a hospital that distinguishes the patient and his or her hospital record from all others in that institution
02. Date of birth Month, day and year of birth. Capture of the full four-digit year of birth is recommended 03. Sex Male or female
An example of UHDDS Data Elements
(Johns, 2002)
○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○
“A data dictionary contains a set of core uniform
definitions and data items relating to the full range of
health services and a range of population parameters
”
(AIHW, 2001).
Data dictionaryA data dictionary is a set of common standards for data collection.
It is used to promote uniformity, availability, reliability, validity, consistency and completeness of data at all levels of health care service. The data dictionary is useful especially in promoting national standard definitions and should be readily available to all individuals and organizations involved in the generation, use and/or development of health and health service information.
The National Health Data Dictionary, Version 10, produced by the Australian Institute of Health and Welfare (AIHW), Canberra, is used extensively by health care services in Australia to ensure the production of standardized data collections. Data elements are based on the ISO/IEC Standards 11179 Specification and Standardization of Data Elements. The dictionary is used to help ensure that data elements are collected uniformly from all health care facilities to improve the quality of information for community discussion and public policy debate on health issues in Australia(AIHW, 2001).
Data dictionaries can also exist at state or province level and should be consistent with the appropriate national standards.
16
Health Outcome
Identifying and definitional attributes :
Data element type : Data element concept
Definition : A change in the health of an individual, or group of people or a population, which is wholly or partially attributable to an intervention or a series of interventions.
Context : Admitted patient and non-admitted patient health care.
A sample entry in the data dictionary
Some countries also produce standards on documentation associated with Accreditation Guidelines for Health Care Facilities, including hospitals and community health care centres and clinics. For example, medical/health record documentation determined by the Joint Commission on Accreditation of Healthcare Organizations (JCAHO, 1995) in the United States of America include standards such as:
IM.7.5.1 Records of emergency visits contain time and means of arrival.
IM.7.5.2 Records of emergency visits contain conclusions at end of treatment, including final disposition, condition and instructions for follow-up.
Example of standards produced by JCAHO
An example of a standard related to medical records published in the First Edition of the Joint Commission International Accreditation Standards for Hospitals (JCIASH, 2000) is as follows:
The clinical record contains sufficient information to identify the patient, support the diagnosis, justify the treatment, document the course and results of treatment, and promote continuity of care among health care providers (MOI.2.1 JCIASH).
Measurable elements of MOI.2.1 include:
1. patient clinical records contain adequate information to identify the patient; 2. patient clinical records contain adequate information to support the diagnosis; 3. patient clinical records contain adequate information to justify the care and treatment;
4. patient clinical records contain adequate information to document the course and results of treatment; 5. patient clinical records promote continuity of care; and
6. the specific content of patient clinical records has been determined by the organization.
The intent of this standard is that the clinical record of each patient needs to present sufficient information to support the diagnosis, justify the treatment provided, and document the course and results of treatment. A standardized format and content of a patient’s clinical record helps promote the integration and continuity of care among the various providers of care to the patient.
Standard MOI.2.1
The Australian Council on Healthcare Standards also provides a guide to health care facility accreditation in which standards for medical records are included. An example of a standard is:
17
Patient/consumer care, management of services, education and research are facilitated by the collection and aggregation of data.
Criteria – data management
4.2.1 Relevant, accurate, quantitative and qualitative data are collected and used in a timely and efficient manner for delivery of patient/consumer care and management of services.
Information created by the organization needs to be timely, accurate, clear, concise and presented in a way that is appropriate to the users’ needs. Excessive quantities or poorly constructed information can deter a person from reading or using it. Regular review of the organization’s reporting systems can assist in identifying how well the information systems meet users’ needs and whether any action is required to improve the system. Those responsible for providing information can monitor their
performance through measuring production delays, rework and frequency of inaccurate information. Feedback from staff may also provide suggestions for improving the systems for managing information.
4.2.2 The collection of data and reporting of information comply with professional and Australian standards and statutory requirements.
Ensuring consistency is an important part of the data collection process. Data consistency is promoted by complying with protocols, definitions and standards. It is also promoted by the use of standard definitions, symbols and abbreviations throughout the organization. Periodic evaluation of data consistency and quality can help ensure the data collected by the organization and its services are appropriate and contribute to management and patient care.
4.2.3 Indexing of data facilitates information retrieval.
Effective indexing of data assists the storage and retrieval of information.
4.2.4 Reference and research information is collected and managed for use by staff and patients/consumers in achieving the organization’s goals.
Clinical records
4.2.5 The organization uniquely identifies all patients/consumers, including newborn infants. The organization needs to consider the efficiency and effectiveness of its system of uniquely identifying patients/consumers.
Other
Other segments of this standard include:
4.2.6 Every patient/consumer has a sufficiently detailed record to assist with the continuity of care, education, research, evaluation and medico-legal and statutory requirements.
4.2.7 Care providers document details in the clinical record. All entries are legible, dated and signed with designation.
4.2.8 In circumstances where part of the clinical record is available electronically, appropriate mechanisms exist to alert staff to the existence of these data.
4.2.9 Clinical record data are coded and indexed to ensure the timely production of quality patient/consumer care information.
Standard 4.2 Data collection, aggregation and use
Benchmarking
Another method of setting standards is benchmarking, which may be defined as “the process of systematically searching for and incorporating international best practice into an organization” (AIHW, 1996). A benchmark may be achieved by comparing an organization, a standard or peer groups (Johns, 2002). It is a quality improvement technique used by one facility to compare its practices with that of another with superior performance by reviewing the process that has proven effective (Davis and LaCour, 2002). Benchmarking, however, may not always be achievable. Some problems that may impede benchmarking include agreement on a benchmark, finding partners to participate and training staff.
18
Leadership in data quality
Many health care administrators already recognize that quality improvement is the way to add value to the services offered and that the dissemination of quality data is the only way to demonstrate that value to health care authorities and the community.
■ Health care administrators/managers should be leaders in the move to improve the quality of data collected in the health care facility as they are responsible for the overall management of the facility and the quality of the information produced.
■ Senior doctors should take the lead in ensuring data quality by taking time to ensure the more junior doctors record clinical data accurately and in a timely manner. Doctors should play an important role in maintaining data quality and should understand the need for accurate and timely data in the care of patients.
■ Senior staff in departments such as laboratory, radiology and pathology and
community nurses, allied health professionals such as physical therapists, occupational therapists and social workers, should be responsible for the provision of accurate and timely reporting and better checks on the quality of the content of reports.
■ Senior medical record staff should ensure that the medical/health record is complete, available and accessible when needed. They should also take responsibility for accurate and timely abstracted data and statistical reports. In a hospital setting, they should ensure that discharge summaries are accurately prepared and that diagnoses and procedures are coded accurately and within a specified time frame. A major part of data quality emanates from the medical record department through the timely preparation of the medical record, quality checks for accuracy and completeness of recorded data, and the availability of the medical/health record at all times when needed for patient care.
Routine data quality monitoring
With standards in place, procedures relating to data collection and monitoring data quality should be carried out on a routine basis.
Two monitoring procedures for inpatients that have been undertaken for many years in some countries are quantitative and qualitative analysis of medical records.
Quantitative analysis of medical records
To evaluate the quality of documentation and, subsequently, patient care, the medical record must be complete. In a quantitative analysis, medical records should be reviewed to check that all documentation has been included (Huffman, 1963), for example:
■ Patient identification is accurate and all details are complete.
■ The history, physical examination, all progress and nursing notes are present, and all relevant reports such as pathology, X-rays, etc., are included.
19
■ If the patient had surgery an operation and anaesthetic report is present.
■ All entries are signed and dated.
In other words, all relevant documentation must be present and authenticated.
In the assembly of medical records, the required standard
is that all medical records are assembled within 24 hours
of patient discharge with 100% accuracy.
When a manual system is used in a hospital, it is often difficult to verify data unless certain steps are built in. These could include routinely performing quantitative analyses on all medical records of discharged patients with verification by a second person. However, this would affect efficiency, as it would delay the processing of the medical record. Alternatively, a regular review could be conducted. The thorough training of staff in the majority of situations, however, would assist in eliminating many errors and improve the data collection.
The medical record department personnel can verify the completeness of the medical record by undertaking a formal medical record review using a random sampling method on a regular basis.
The implementation of such a system would enable staff to review data collected, identify deficient areas, provide a system of notification to individuals responsible for the collection of data, and provide follow-up for correcting deficiencies (Abdelhak et al., 1996).
A sample medical record review form can be found in Appendix 1.
Again, this procedure could be used in a non-inpatient environment with minor adjustments to the form, so that data collected by health professionals and important statistical data can be verified. In a computer-based patient record, most data are captured electronically with built-in checks and edits, as previously discussed, which should reduce the margin of error.
Qualitative analysis of medical records
In a qualitative analysis of medical records, the information pertaining to patient care is reviewed for accuracy, validity and timeliness (Huffman, 1990). This includes:
■ reviewing medical records to ensure that all clinically pertinent data have been accurately recorded; and
■ checking the front sheet to ensure that the patient’s diagnosis and treatment have been recorded and are supported by documentation in the body of the medical record.
The quality of health care is measured by the quality of
the data in the medical record.
In an ideal situation, a staff member trained in quality assessment should perform a qualitative analysis on every medical record of discharged patients. This procedure takes a significant amount of
20
time and in most situations, there are insufficient staff with the time available to complete the job effectively. The administrator or manager of a health care facility is generally responsible for determining what type of record to review and how often they should be reviewed.
To be an effective tool in data quality, a qualitative
analysis should be performed at three- to four-month
intervals on a random sample of patient records. It
could be on 30 or 5% of records of discharged patients,
or on a specific diagnosis or procedure performed.
In addition to the above, other procedures that could be implemented to ensure data quality are listed below.
Data entry checks
In a manual system, steps to check the entry of data in the medical/health record should be taken at the point of entry.
■ Check the collection of demographic data by clerical staff in the admission office, emergency department and outpatient department reception, and health centre and clinic reception/ registration area prior to the provision of health care services. The accuracy of this data is crucial for the identification of the patient during the present visit as well as their admission or future encounters with the health care service. Regular checks should be in place to prevent incorrect data being entered.
■ Doctors and other health care professionals at all levels of health care should check the accuracy of the data for which he or she is responsible. In most medical/health records, data are recorded by a variety of persons, all of whom are responsible for the accuracy of his or her documentation. That is, the responsibility for accurate and timely data entry rests with the professionals involved and regular checks should be undertaken.
■ When the medical/health record is returned to the medical record department (or the file room after attendance at a centre or outlying post) after discharge or death or outpatient attendance, staff are responsible for checking for completion before filing. They are responsible for monitoring the quality of the data and ensuring that health information generated from the medical record is timely, complete, accurate and accessible. The person responsible for the health record services, regardless of the type and level of health care, must manage those services in a manner that promotes quality information.
Checks on the quality of abstracted data
If a qualitative analysis is not undertaken on the complete medical/health record, a data quality check should be carried out on abstracted data. For most inpatient health care services, an abstract, that is
21
the abstraction of information from a document to create a summary, is prepared at the end of a patient’s hospital stay by the attending doctor. It is necessary to ensure the quality of the abstract information, for accuracy, validity, completeness and timeliness.
■ A staff member, other than the initial clerk should routinely check the abstracts. To do this, the patient’s record is retrieved from the database, or manually if a non-computerized record system is maintained, and the data elements are verified. In most cases, random samples are undertaken, errors noted are corrected and documented, and the staff member responsible for the original abstract re-trained.
■ In a paper record, if a doctor forgets to sign an order, the order is not authenticated and cannot be carried out until the doctor signs.
■ In a computer-based or electronic patient record, the authentication of an order is captured by a key word or code, and entered by the doctor when he or she has completed the order.
There are various methods of capturing authentication data — they are referred to as electronic signatures. An electronic signature does not capture a person’s actual signature into the computer, what happens is that the computer recognizes a unique code that only the author has in his or her possession. (Davis and LaCour, 2002) The computer can be programmed to reject orders that do not contain appropriate authentication. ”The computer would look for both the existence of the authentication [for data completeness] and the correct authentication [for data validity]” (Davis and LaCour, 2002).
■ This procedure can also be used when abstracting data on non-inpatient services offered at a clinic or aid post. The type of data extracted in this situation is usually of a statistical nature and should be checked for quality on a regular basis.
Audit of accessibility of medical/health records
To help assess the accessibility and the quality of the medical/health record services, study programmes should be instituted. For example, the medical record retrieval process should be studied using pre-determined criteria and data collection form (see Appendix 2).
Medical/health record audit
Similar to a medical/health record review is a medical/health record audit, which is also a retrospective review of selected medical/health records or data documents to evaluate the quality of care or services provided compared with predetermined standards. To validly assess the completeness, accuracy, consistency, and legality of the medical record, an evaluation of the adequacy of medical record content can be conducted. A sample data collection form can be found in Appendix 3.
Some steps identified by Jones (1993) following an audit of hospital medical records are listed below.
■ The health facility administration and senior doctors should be asked to seek improvement in medical record documentation by assisting with development and design strategies to enhance data collection formats.
22
■ Provision should be made for the allocation of sufficient resources to adequately monitor data quality.
■ Support should be obtained from the administration to work with clinical departments and senior clinicians to examine strategies for the provision of adequate patient data.
■ A comprehensive training programme on documentation practices for junior doctors should be developed with the support of the hospital administration and senior medical staff.
■ Continuous auditing of documentation practices should be carried out and findings monitored regularly.
■ A multi-professional forum should be set up to address documentation and other issues and consider using a total quality management approach for improving the quality of data.
Development of an on-going quality assessment plan
If a health care facility or ministry of health are serious about data quality, they need to develop a plan aimed at improving and maintaining the quality of data and the information generated from that data.
To develop a quality plan, several structural components need to be in place. These include:
■ commitment by top-level management to support the programme, which would involve the appointment of a quality co-ordinator with adequate clerical support; and
■ staff responsible for quality control should be involved and deal with quality reports properly by reading and acting upon recommendations in a timely manner (Schofield, 1994).
Remember that data validity is a measure of data
quality. Only valid data entered within a legitimate
range of values for that type of data should be collected.
For example, in a computer system, when collecting dates
of birth, the computer is looking for an eight-digit
number — if more or less, the entry is invalid.
To develop a data quality assessment plan, whether in a hospital, health centre, clinic or aid post, the administration should take certain initial steps, which include:
■ assign responsibility – a specific staff member should be assigned to audit aspects of documentation contained in the patient’s medical/health record;
■ identify important aspects of data collection – such as accuracy and validity, reliability, completeness and timeliness;
■ determine indicators of data quality for each documentation component;
■ set a threshold – that is, determine an acceptable error rate;
■ develop an organized method for collecting data according to quality indicators previously developed;
23
■ assess actions taken to improve documentation; and
■ communicate the results of the review/audit to those affected.
Quality assessment should be undertaken to ensure health information management functions are working effectively within the standards previously determined.
Performance improvement techniques
Along with a continuous quality assessment plan, steps should be taken to institute performance or quality improvement. This is a process by which a health care facility reviews its services to ensure quality (Davis and LaCour, 2002).
■ Staff responsible for health care services should be encouraged to not only meet a certain standard but should also seek to improve their performance.
■ Performance improvement should not only include the staff of medical record services but also the entire staff of the facility and should be multi-disciplinary. To improve data quality, all persons involved in completing, checking and using data should be involved in insuring that the data are correct, valid, timely and relevant.
■ Employees should use teamwork to improve a process. By instituting performance improvement in data collection, the outcome – patient care – will ultimately be improved.
A popular method of performance improvement used in the United States of America is the “Plan, Do, Check and Act” (PDCA) method developed by Walter Shewhart (Abdelhak et al., 1996). The steps are:
■ the plan phase consists of data collection and analysis to propose a solution to a specific problem;
■ the do or implementation phase tests the proposed solution;
■ the check phase monitors the effectiveness of the solution over a period of time; and
■ the act phase formalizes the changes that have proven effective in the “do” and “check” stages.
The important point to remember is that a process is being improved. Concentration should be placed on the process and not on the employee performing the job. To improve the process, ALL persons involved must be part of the team. As outlined by Abdelhak et al. (1996), the PDCA plan is as follows:
Plan
■ Co-ordinate a team
■ Investigate the problem – gather facts
■ Discuss potential solutions
24
Do
■ Test the plan of action
■ Educate employees on the new process
■ Pilot the new process
Check
■ Monitor the new process during the pilot study
■ Did the plan of action work the way the team intended?
■ Make necessary adjustments and continue the pilot
Act (When certain the process is an improvement)
■ Change the policy
■ Educate and train all affected employees
■ Implement the new process
By carrying out a data quality improvement plan, the review of medical/health records provides information to patient care committees, doctors, administrators and outside agencies.
Other steps to assist with data quality
improvement
■ Performance indicators can be developed as a guide to monitor and evaluate the quality and appropriateness of care. They are reliable and could be used to detect change such as health outcomes.
■ Policies should be defined concerning the facility’s overall position on quality. The policies should reflect a commitment to the highest standard of care for patients, including accurately and competently documented demographic and clinical information, and an opportunity for input to the programme from all staff, with full support from the administration.
■ A quality review committee should be established and charged with the responsibility of overseeing the quality activities programme.
■ A quality coordinator should be responsible for the day-to-day co-ordination of the programme. This person must understand quality and its implementation and must be an effective communicator with the ability to impart knowledge to others. That is, a resource person who needs to network within and outside the organization.
■ The data collection system needs to be simple and user-friendly.
■ Quality activities need to focus on practice and not individual workers.
■ Confidentiality needs to be maintained in all programmes.
■ Staff education requires that all staff clearly understand what quality means to the facility, how the programme is managed, what is expected of staff and what they can expect to achieve.
25
Limitations in overcoming problems
related to data quality
Data quality can be hampered by a number of issues, including the following.
■ Lack of uniformity of data – without predetermined standards and uniform data sets, problems relating to the quality of health care data are difficult to solve.
■ Poorly designed data collection forms – if forms are not well designed, the collection of data could be affected, resulting in poor quality data.
■ Limitations to doctors’ capacity to communicate – some doctors find it difficult to record data in a clear and concise manner, resulting in poor information. They also often use non-standard abbreviations and are “too busy” to complete medical records once the patient has been discharged from the facility or does not require further treatment. Limited education in documentation requirements of medical staff is a major factor in poor data quality.
■ Limitations to information transfer from different parts of the facility – sometimes information being transferred from the laboratory to the ward or a clinic does not contain the correct patient’s name and medical/health record number. Such errors make it difficult to ensure that all data pertaining to an individual patient are filed in that patient’s medical record. The transfer of information from one department to another or from a hospital to a clinic or aid post is often slow or information is lost in transmission.
■ Limited education of processing staff –the processing of medical/health records requires staff that can understand the need for accuracy and completeness. If they are not properly trained, the production of quality data is threatened.
■ Lack of planning by administrative staff to ensure data quality control programmes are in place. All data collection and abstracting staff should be properly trained; and doctors should be educated in the requirements for accurate and timely documentation of patient care details.
■ No single record – a problem of quality arises if more than one medical record is kept on each patient. Some facility staff, such as in cardiology, oncology and social work, insist on keeping their own records, thus limiting the overall collection of meaningful data about an individual patient.
■ Data discrepancies –arisingwhen errorsoccur at the point of collection and plans are not in place to check the entry and verify the data.
Summary
For the quality of data in health care delivery to be improved, administrators need to take the following steps:
26
■ Policies should be in place to cover all staff on data collection requirements. If the administration is committed to data quality, it needs to provide resources to establish a formal monitoring programme to ensure quality control. For example, if carrying out a re-abstracting study, a baseline measure of data quality needs to be established.
■ Education – programmes should be offered for medical, nursing and other health care staff as well as medical record staff on the importance of careful collection of original source data. Education could also be aided by the creation of a videotape on data quality for staff, including doctors, and the development of simple data quality guidelines.
■ Tools – sufficient equipment and technology support should be available for all staff. Access to a computer for patient identification and other data collection requirements would provide staff with the ability to edit work as it is being done and greatly reduce errors.
To ensure documentation meets the required standards previously determined, quality assessment, record review and improvement procedures should be developed. Poor quality data are a major hindrance to planning and decision-making, therefore, data quality is an important concern for health care delivery at anytime in any part of the world.
References
Abdelhak M, Grostick S, Hankin MA, Jacobs E. Health Information: Management of a Strategic Resource.
Philadelphia, WB Saunders, 1996.
Australian Council on Health Care Standards. The EQUIP Guide: Standards and Guidelines for the Evaluation and Quality Improvement Programme (2nd edition). Sydney, ACHS, 1998.
Australian Institute of Health and Welfare. National Health Data Dictionary: Version 10. Canberra, AIHW, 183, 2001.
Davis N, LaCour M. Introduction to Information Technology. Philadelphia, WB Saunders, 2002.
Donabedian A. Quality Assessment and Assurance: Unity of Purpose, Diversity of Means. Inquiry: Vol. 25. 173–195, 1988.
Huffman E. Manual for Medical Record Librarians. Chicago, Physician’s Record Company, 1963. Huffman E. Medical Record Management (9th edition) (Finnegan R, Amatayakul M, eds.). Berwyn, Ill.,
Physician’s Record Company, 1990.
Johns ML. Health Information Management Technology: An Applied Approach. Chicago, American Health Information Management Association, 2002.
Jones L. The Quality of Clinical Documentation and Subsequent Effect on DRG Assignment: A Report on the Findings of the DRG Documentation Study. Canberra, Case Mix Development Programme, Commonwealth Department of Health, Housing and Community Services, 1993.
27
Joint Commission on Accreditation of Healthcare Organizations. Comprehensive Accreditation Manual for Hospitals. Oakbrook Terrace. Ill., JCAHO, 1995.
Joint Commission on Accreditation of Healthcare Organizations: First Edition of the Joint Commission International Accreditation Standards for Hospitals, 2000.
Schofield J. Developing a Quality Plan. Health Care Quality Activities in Practice in Australia, 4th Edition.
28
Quality in data
collection
3
Building quality into the process of data collection right from the beginning
is one of the major foundations of the overall move to data quality control.
The main source of health care data is the patient’s medical/health record, which should contain essential data used in the health care decision-making process. It is the “who, what, when, where, why and how” of patient care (Abdelhak et al., 1996). A patient’s medical/health record should provide accurate information on:
■ who the patient is and who provided health care;
■ what services were provided;
■ when and where the services were provided;
■ why the services were provided;
■ how effective the services were; and
■ what the outcome was of care and treatment.
Another important collection of data by hospitals is the bed census, which is collected daily and processed monthly and annually to produce statistics on the utilization of hospital services.
Health care facilities, regardless of type or size, collect data processed as statistics for specific external reporting to meet the needs of health care authorities, local and national governments, and funding organizations.
Statistical data are collected from clinics and outposts on the number, type and age of patients attending the facility, the types of conditions treated, tests conducted and referrals.
The collection of the original source data must have in-built procedures to ensure that they meet the required quality standards and also the requirements of the patient, health care provider, health care facility and government.
T
he collection
of the
original source
data must have
in-built
procedures to
ensure that they
meet the
required quality
standards.
29
Type of data collected and how collected
The collection of accurate and timely data has a significant impact on the efficiency and effectiveness of the decision-making process in health care.
Data quality should
start at this point.
The type of primary health care data collected about a patient in either a manual medical/ health record or computer-based or electronic patient record at all levels of health care fall into two major categories (Abdelhak et al., 1996).
Administrative data
and clinical data.
Administrative data
Administrative data – can be divided into four sections: demographic, legal, financial and provider (Abdelhak et al., 1996).
■ Demographic and socioeconomic data – personal data elements about a patient. It is essential that these data are accurately recorded at the initial point of contact of the patient and are sufficient to identify an individual patient. Demographic data include the name of the patient, sex, date of birth, place of birth, the patient’s permanent address and a unique personal identifier. This is usually a number, whether for national identification, social security or another piece of information that will uniquely identify the patient.
■ Legal data – should include a signed consent by the patient for health care services at a hospital or primary health centre and a signed and dated authorization for the release of information. A signed and dated authorization for specific procedures and treatment is also required. In addition, in the event of the patient’s death, an authorization for an autopsy signed and dated by the patient’s next of kin should be in the medical record.
■ Financial d